The aggressive technology scaling imposes severe reliability challenges on present-day VLSI designs, e.g., the probabilistic behaviors of nano-scale circuitry due to process variations. Logic gates and memory cells also turn to be highly vulnerable to radiation-induced soft-errors due to the limited electrical charge stored on the internal capacitance. In this project, we propose a comprehensive solution set combating the statistical properties and intermittent failures incurred by the technology scaling in computing systems. In contrast to many existing approaches that focus on circuit and structural enhancements, our proposal offers an adaptive information processing VLSI system with significant revolutions across the levels of device, circuit, architecture and algorithm. The design starts with the adoption of emerging memristor technology which is intrinsically resilient to soft errors and friendly to technology scaling. A robust computing engine built on high-density memristor crossbar array with CMOS/memristor hybrid mixed-signal peripheral circuitry is then proposed for intensive computations. Novel computation algorithms with online programming and noise filtering will be developed to accommodate the stochastic behaviors of memristors incurred by process variations and environmental fluctuations, offering high computation parallelism, high power efficiency and high system reliability. The proposed design aims at a broad spectrum of military applications with rich matrix operations and learning process. A successful application example inspired by our collaboration with AFRL – document image processing, is presented.
|Senior Personnel:||Prof. Hai (Helen) Li (PI)|
|Sponsors:||Defense Advanced Research Projects Agency|
|Team Members:||Miao Hu, Chenchen Liu, Bonan Yan, and Chaofei Yang|
|Collaborators:||Dr. Qing Wu, U.S. Air Force Research Lab
Dr. Jianhua Yang, HP Labs
M. Hu, H. Li, Y. Chen, Q. Wu, G. Rose, and R. Linderman, “Memristor Crossbar Based Neuromorphic Computing System: A Case Study,” IEEE Transactions on Neural Networks and Learning Systems (TNNLS), vol. 25, no 10, pp. 1864-1878, Oct. 2014. DOI: 10.1109/TNNLS.2013.2296777.